CLT ASIS knitr_report

Gaonyalelwe Maribe

System doc

The ASIS reporting application is written using the R Shiny web development framework. The project includes 9 R header scripts and one rmarkdown written using knitr.

This is a private repository containing the ASIS -CTL project for generating impact reports using student attendance and performance data. This repository is maintained by xdelta analytics.

Software Requirements

You will need to install the following software on your system to edit or configure the r scripts, markdown files, the yaml files, dockerfile and header files. For testing and deployment. For the docker image, these requirements are already listed in the dockerfile.

• ubuntu 18.04 os
  • docker.io, pandoc, gdebi-core, pandoc, pandoc-citeproc, libcurl4-gnutls-dev, libcairo2-dev/unstable, libxt-dev, libssl-dev, libxml2-dev
• r-base:3.5.0
• r-studio: latest
• r libraries to install
  • ‘knitr’,‘shiny’, ‘shinydashboard’,‘ppcor’, ‘dplyr’,‘rmarkdown’,‘DT’,‘htmlwidgets’, ‘httpuv’,‘shinyalert’,‘Hmisc’, ‘plotly’,‘kableExtra’,‘shinyjs’, ‘kableExtra’, ‘phantomjs’
• GitHub and Dockerhub

Set-Up Configurations

Environments

• Rprofile: Ensure that shiny is listening on localhost with port 3838

local({
   options(shiny.port = 3838, shiny.host = "0.0.0.0")
})

• knitr

knit_hook$set(set_envir = function(before, options, envir){
  if(before)
    set.environment(options$set_envir)
  else
    set.envir(globalenv())
})

options(tinytex.verbose = TRUE)
knitr::opts_chunk$set(echo = TRUE)
knitr::opts_chunk$set(echo = TRUE, fig.align="center")
knitr::opts_chunk$set(echo = TRUE, warning=FALSE)

Accounts to setup:

• GitHub [xdelta will provide access to the ASIS repo]
• Dockerhub [xdelta will provide access to the ASIS repo]
• UFS access to the asis_server [Developers on the project will have to apply to ICT for access]

Set-up the container manager

The system uses the shinyproxy to manage containers. The configuration file for shinyProxy can be found on the server $HOME directory in ~/shinyproxy/target/application.yml. Documentation for this is available on the shinyproxy page.

The production docker image can be built on prem. (documentation on this can be found on the docker website). The image can then be pushed to dockerhub:

An account will first need to be created. Then login from your command line interface (CLI):

docker login --username=yourhubusername --email=youremail@company.com

Check if the docker image exists:

docker images

Tag the image:

docker tag b************ yourhubusername/knitr_report:asisV

And you should be ready to push:

docker push yourhubusername/knitr_report:asisV

Data uploading script

data_collection1.R contains the data_collection_ui1 function which is called by the server script to paste the ui of the first page for the user to upload the attendance data file. This function is only called if the complete_data.RData R object is not found in the project directory. The server variable e is then set globally to zero, indicating that the current ui being used is data_collection_ui1.

data_collection2.R contains the data_collection_ui2 function which is called by the server script to paste the ui of the second page for the user to upload the performance data file. The server variable e is then set globally to 1, indicating that the current ui being used is data_collection_ui2.

Data aggregation script

The data aggregation script data_compile.R uses the two uploaded ‘delimiter corrected’ data sets attendance and performance. These two files are stored on the $HOME directory. The data is then aggregated into a dataframe called GroupedData and stored in a file called “complete_data.RData”. This file is attached to every docker instance volume and should be purged everytime your docker instance is stopped.

Reporting

The server.R file is well commented, it uses the reporting headers from the Decision_Function.R script which imports the GroupedData and runs decisions on the subset modules (parameters from the reporting_ui). The server.R file then writes all the report text into a file called our_info and prints all report images into the $images directory attached to the docker volume.

User doc

Data import

When prompted, the user should upload the attendance data from the CTL attendance dashboard. This data should look like the attached figure:

Once this data is uploaded, the user has to chose a delimiter that separates the columns of the data correctly into a table. You will be given some instructions at the bottom of the page, these instructions will guide you when requesting for data from DIRAP to ensure you have the correct performance data.

Once you’re satisfied proceed to pressing the next button on the bottom left of the sidebar. The user will then be prompted to upload the performance data. Follow again the same process in separating the columns correctly into a table and press done. The files will be proced by the data_compile.R script and the resulting GroupedData stored into an RData object.

By now your user interface will be showing a slidebar with some parameters that you can change as you please and a main panel showing some prelim results on a modular level. After selecting your parameters, generate your report and download a word version to edit as you please.